Improved point estimation for inverse gamma regression models

This paper develops a bias correction scheme for reparametrized inverse gamma regression models with varying precision [Bourguignon M, Gallardo DI. Reparametrized precision. Stat Neerl. 2020;74(4):611–627], which is tailored to situations where the response variable has an asymmetrical shape on positive real line. In particular, we discuss maximum-likelihood estimation model parameters and derive closed-form expressions first-order of estimators. The derived are simple only require definition few matrices. enables us obtain corrected estimators that approximately unbiased. We conduct extensive Monte Carlo simulation study evaluate performance proposed Finally, apply results obtained in three real-world datasets. contains Supplementary Material.